Saeed Ghorbani

saeed.jpg

Research Scientist

Amazon Games

Toronto, Canada

Contact

I am a Research Scientist at Amazon Games, specializing in machine learning, computer vision, computer graphics, and computer animation. Previously, I worked as an R&D Scientist at Ubisoft La Forge, where I contributed to cutting-edge research in motion synthesis and gesture generation.

I earned my PhD from York University, where I was a trainee with both the VISTA and CVR programs. My research focuses on developing novel deep probabilistic models for realistic human motion modeling, with applications in character animation, pose estimation, and gesture synthesis.

News

Jul 19, 2024

Our paper on human pose estimation with cross-view and temporal cues has been accepted to ECCV 2024.

May 15, 2023 ZeroEGGS Featured in Ubisoft La Forge Blog
Mar 13, 2023

Zero-shot Example-based Gesture Generation from Speech is published at Computer Graphics Forum. Link to the paper

Jan 15, 2023 Joined Amazon Games as Research Scientist
Nov 28, 2022

ZEGGS is featured in Two Minute Papers channel.

Selected publications

  1. ECCV
    UPose3D: Uncertainty-Aware 3D Human Pose Estimation with Cross-View and Temporal Cues
    Vandad Davoodnia, Saeed GhorbaniMarc-André Carbonneau, Alexandre Messier, and Ali Etemad
    European Conference on Computer Vision (ECCV) 2024
  2. ECCVW
    SkelFormer: Markerless 3D Pose and Shape Estimation using Skeletal Transformers
    Vandad Davoodnia, Saeed Ghorbani, Alexandre Messier, and Ali Etemad
    ECCVW Workshop on Video Games 2024
  3. CGF
    ZeroEGGS: Zero-shot Example-based Gesture Generation from Speech
    Saeed Ghorbani, Ylva Ferstl, Daniel HoldenNikolaus F. Troje, and Marc-André Carbonneau
    Computer Graphics Forum 2023
  4. CGF
    Probabilistic Character Motion Synthesis using a Hierarchical Deep Latent Variable Model
    Computer Graphics Forum (Symposium on Computer Aanimation) 2020
  5. PLOS ONE
    MoVi: A large multi-purpose human motion and video dataset
    Saeed Ghorbani, Kimia Mahdaviani, Anne Thaler, Konrad Kording, Douglas James Cook, Gunnar Blohm, and Nikolaus F. Troje
    Plos one 2021
  6. CGI
    Auto-labelling of markers in optical motion capture by permutation learning
    Saeed GhorbaniAli Etemad, and Nikolaus F. Troje
    In Computer Graphics International 2019